GPU-Accelerated Discrete Wavelet Transform for Images

Abstract:

Discrete Wavelet Transform (DWT) has been brought into wide use in image processing, but it cant fit the demand of the hugeimage data because the time of computing is vast. The GPU is an attractive platform for a broad fieldof applications,which remains asignificanthigharithmetic processingcapability. Therefore itcan beusedasa powerful accelerator without extra cost.CUDA(computeunifieddevicearchitecture) providesahardwareandsoftwareenvironment touse the GPU to accelerate the DWT for images. In this paper, we use the NVIDIA GeForce GT 650M that complies with the CUDA to improvethe execution time of theDiscrete Wavelet Transformfor images. TheresultofexperimentsindicatesthattheCUDAtechnology hastheadvantagesof parallel processingandtheefficiencyofimagetransform isimprovedgreatly. Whats more, it performs better on the larger size image (the max speedup is 15.9).

• Periodical: Advanced Materials Research (Volumes 718-720)
• Pages: 2086-2091
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.718-720.2086
• Online since: July 2013
• Authors: Yu Min Chen, Fei Zeng, Jing Yang Wu, Qiao Wan, Zhi Jun Su
• Keywords: CUDA, Discrete Wavelet Transform (DWT), Graphic Processing Unit (GPU)

References

[1] A. S. Lewis and G. Knowles, Image compression using the 2-d wavelettransform, IEEE Transactions on Image Processing, 1992, vol. 1 of2, pp.244-256.

[2] NVIDIA Tutorial at PDP08,CUDA: A new architecture for computingon the GPU,2008.

[3] J.D. Owens, S. Sengupta, and D. Horn, Assessment of Graphic Processing Units (GPUs) for Department of Defense (DoD) Digital Signal Processing (DSP) Applications, Technical Report, ECE-CE-2005-3, Computer Engineering Research Laboratory, University ofCalifornia, Davis, October 2005.

[4] M.Hopf, and T.Ertl, Hardware Based Wavelet Transformations, Vision,Modeling and Visualisation'99, IEEE and GI Infix Press, November 1999, pp.317-328.

[5] G.Bernabe, CUDA and OpenCLImplementations of 3D FastWavelet Transform,Circuits and Systems (LASCAS), IEEE Third Latin American Symposium, 2012, pp.1-4.

DOI: 10.1109/lascas.2012.6180318

[6] M.Frydrych and J.Parkkinen, Wavelets and Natural Image StatisticsProceedings of the 10thScandinavian Conference on Image Analysis, 1997, pp.13-18.

[7] J. M. Shapiro,Embedded Image Coding Using ZerotreesofWaveletsCoefficients, IEEE Transactions on SignalProcessing, December 1993, pp.3445-3462.

[8] D.Taubman, High performance scalable image compression with ImageProcessing, IEEE Transactions,2000, pp.1158-1170.

DOI: 10.1109/83.847830

[9] V.S. Petrovic, C.S. Xydeas,Gradient-Based Multiresolution Image Fusion, IEEE Trans. Image Process, 13(2), 2004, pp.228-237.

DOI: 10.1109/tip.2004.823821

[10] Information onhttp://geco.mines.edu/tesla/cuda_tutorial_mio/index.html.

[11] Information on http://developer.download.nvidia.com/compute/cuda/1_0/NVIDIA_CUDA_Programming_Guide_1.0.pdf.

[12] J.Nickolls,I.Buck, M.Garland, K.Skadron,Scalable Parallel Programmingwith CUDA,ACM 6(2),2008, pp.40-53.

DOI: 10.1145/1401132.1401152

[13] Han-Wei Shen Antonio Garcia, GPU -based 3d wavelet reconstructionwith tileboarding, in The Vis ual Computer, 2 1(8C10):, 2005, pp.755-763.

DOI: 10.1007/s00371-005-0332-0

[14] M. J. Nadenau, M.KuntJ.Reichel, G. Menegaz, GPU acceleration of2d-dwt image compression in matlab with cuda, in Second UKSIMEuropean Symposium on Computer Modeling and Simulation , 2008, pp.274-277.

DOI: 10.1109/ems.2008.43